Transform Coding Patents (Class 382/248)
  • Patent number: 11172233
    Abstract: Apparatus, systems and methods for adaptively reducing blocking artifacts in block-coded video are disclosed. In one implementation, a system includes processing logic at least capable of deblock filtering at least a portion of a line of video data based, at least in part, on edge information and texture information to generate at least a portion of a line of deblocked video data, and an image data output device responsive to the processing logic.
    Type: Grant
    Filed: September 29, 2020
    Date of Patent: November 9, 2021
    Assignee: Intel Corporation
    Inventors: Jorge E. Caviedes, Mahesh M. Subedar, Khasim S. Dudekula
  • Patent number: 11143729
    Abstract: The present invention improves the quality of the magnitude and phase images produced in medical imaging, in particular in the case of multi-antenna MRI. The invention proposes to generate (28) such an image (I) by summing the complex image data (pj,{tilde over (p)}j) obtained from different antennas by weighting these data using only the diagonal elements (Rj,j) of an antenna noise covariance matrix or its inverse or pseudo-inverse matrix (Rj,j?1). A reference antenna (Ref) may be determined (24), so as to be able to replace (26), in each of these datum, a phase component specific to the acquisition antenna by a reference phase component. The reference antenna is preferably a virtual antenna formed by linear combination of the antennas of the MRI system. If significant improvements are obtained in the phase image resulting from the summation, a very clear gain is also surprisingly obtained in the magnitude image.
    Type: Grant
    Filed: October 15, 2018
    Date of Patent: October 12, 2021
    Assignees: INSTITUT DU CERVEAU ET DE LA MOELLE EPINIERE (ICM), CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE (CNRS), INSTITUT NATIONAL DE LA SANTE ET DE LA RECHERCHE MEDICALE (INSERM), ASSISTANCE PUBLIQUE—HOPITAUX DE PARIS (AP-HP), SORBONNE UNIVERSITE
    Inventor: Mathieu Santin
  • Patent number: 11068745
    Abstract: Disruption of computerized face detection includes receiving a source image that contains a representation of a face and computing a perturbation for the source image. The perturbation is specific to the source image and is configured for a target face detector. A perturbed image is then generated by adding the perturbation to the source image and then the perturbed image may be outputted instead of the source image.
    Type: Grant
    Filed: September 14, 2018
    Date of Patent: July 20, 2021
    Assignee: DE-IDENTIFICATION LTD.
    Inventors: Avishek Bose, Parham Aarabi
  • Patent number: 11064219
    Abstract: There is disclosed herein a method of encoding an image file, the method comprising inputting dimension, channel, bit depth and color model data of the image file; forward transforming the image file; entropy coding the image file; inverse transforming the image file; and, reconstructing the image file.
    Type: Grant
    Filed: December 3, 2019
    Date of Patent: July 13, 2021
    Assignee: Cloudinary Ltd.
    Inventor: Jon Sneyers
  • Patent number: 10922882
    Abstract: Embodiments of the systems and methods described herein provide game terrain generation system that can generate height field data from a sketch of graphical inputs from a user via a graphical user interface. The game terrain generation system can use a model, such as a trained neural network, to apply macro and micro topological features on top of the height field data to generate game terrain data. The game terrain generation system can identify boundaries between different styles of terrain and generate transitions between the styles to create a more realistic terrain boundary.
    Type: Grant
    Filed: February 8, 2019
    Date of Patent: February 16, 2021
    Assignee: Electronics Arts Inc.
    Inventors: Han Liu, Yiwei Zhao, Mathieu Guindon, Mohsen Sardari, Harold Henry Chaput, Navid Aghdaie, Kazi Atif-Uz Zaman, Kenneth Alan Moss
  • Patent number: 10917644
    Abstract: In various embodiments, an iterative encoding application encodes a source video sequence. The encoding optimization application generates a set of shot encode points based on a set of encoding points and a first shot sequence included in the source video sequence. Each shot encode point is associated with a different encoded shot sequence. The encoding optimization application performs convex hull operation(s) across the set of shot encode points to generate a first convex hull associated with the first shot sequence. Subsequently, the encoding optimization application generates encoded video sequences based on the first convex hull and a second convex hull associated with a second shot sequence included in the source video sequence. The encoding optimization application computes a new encoding point based on the encoded video sequences and a target value for a first video metric and then generates an optimized encoded video sequence based on the new encoding point.
    Type: Grant
    Filed: February 22, 2018
    Date of Patent: February 9, 2021
    Assignee: NETFLIX, INC.
    Inventor: Ioannis Katsavounidis
  • Patent number: 10911791
    Abstract: In various embodiments, a buffer-based encoding application generates a first convex hull of subsequence encode points based on multiple encoding points and a first subsequence associated with a media title. The buffer-based encoding application then generates a first global convex hull of media encode points based on a transmission buffer constraint, the first convex hull, and a second global convex hull of media encode points. Notably, the second global convex hull is associated with a portion of the media title that occurs before the first subsequence in a playback order for the media title. Subsequently, the subsequence-based encoding application selects a first media encode point included in the first global convex hull based on a media metric and determines a first encoded media sequence based on the selected media encode point. The first encoded media sequence is subsequently streamed to an endpoint device during playback of the media title.
    Type: Grant
    Filed: January 9, 2019
    Date of Patent: February 2, 2021
    Assignee: NETFLIX, INC.
    Inventor: Ioannis Katsavounidis
  • Patent number: 10818083
    Abstract: A depth-first walk of a tree structure is performed to populate nodes of the tree structure with corresponding merged tiles of a merged data pyramid to be generated during the walk from source tiles. At a first visit of a target node, a source tile from a first source data layer corresponding to the target is fetched and stored in memory as a first source-derived tile corresponding to the target. At a last visit of the target, a resampling on source-derived tiles from a second source data layer stored in memory corresponding to children nodes of the target node is performed to generate and store in memory a second source-derived tile corresponding to the target node. Also, at the last visit of the target, at least the first and second source-derived tiles are merged to generate a merged tile of the merged data pyramid corresponding to the target.
    Type: Grant
    Filed: April 30, 2019
    Date of Patent: October 27, 2020
    Assignee: International Business Machines Corporation
    Inventor: Samuel Richard Baskinger
  • Patent number: 10778412
    Abstract: In one embodiment, an apparatus comprises a memory and a processor. The memory is to store visual data associated with a visual representation captured by one or more sensors. The processor is to: obtain the visual data associated with the visual representation captured by the one or more sensors, wherein the visual data comprises uncompressed visual data or compressed visual data; process the visual data using a convolutional neural network (CNN), wherein the CNN comprises a plurality of layers, wherein the plurality of layers comprises a plurality of filters, and wherein the plurality of filters comprises one or more pixel-domain filters to perform processing associated with uncompressed data and one or more compressed-domain filters to perform processing associated with compressed data; and classify the visual data based on an output of the CNN.
    Type: Grant
    Filed: June 29, 2018
    Date of Patent: September 15, 2020
    Assignee: Intel Corporation
    Inventors: Yen-Kuang Chen, Shao-Wen Yang, Ibrahima J. Ndiour, Yiting Liao, Vallabhajosyula S. Somayazulu, Omesh Tickoo, Srenivas Varadarajan
  • Patent number: 10743028
    Abstract: Methods of encoding and decoding for video data are described for encoding or decoding coefficients for a transform unit. In particular, the sign bits for the non-zero coefficients are encoded using sign bit hiding. Two or more sets of coefficients are defined for the transform unit and a sign bit may be hidden for each set, subject to satisfaction of a threshold test. The sets may correspond to coefficient groups that are otherwise used in multi-level significance map encoding and decoding.
    Type: Grant
    Filed: February 12, 2018
    Date of Patent: August 11, 2020
    Assignee: Velos Media, LLC
    Inventors: Jing Wang, Xiang Yu, Dake He
  • Patent number: 10699219
    Abstract: A method for computing a probability that an object comprises a target includes: performing a scan of an area comprising the object, generating points; creating a segment corresponding to the object using the points as segment points, the segment extending from a first segment point to a last segment point, the segment comprising a plurality of the segment points; and applying a metric, computing the probability that the object comprises the target.
    Type: Grant
    Filed: October 11, 2019
    Date of Patent: June 30, 2020
    Assignee: Fetch Robotics, Inc.
    Inventors: Alex Henning, Michael Ferguson, Melonee Wise
  • Patent number: 10645386
    Abstract: An embedded codec (EBC) circuitry includes encoder circuitry to encode an image block by application of a sequential encoding scheme to generate an encoded image block. The encoder circuitry determines a local pixel value of each pixel of the plurality of pixels in the encoded image block, based on a quantization bin size of each pixel in the encoded image block. The encoder circuitry selects a reconstruction point from a plurality of reconstruction points based on an accumulated difference of the local pixel value and an input pixel value of each pixel in the encoded image block. The encoder circuitry allocates a set of signaling bits that indicates the selected reconstruction point in a bit-stream of the encoded image block. The EBC circuitry further includes decoder circuitry to reconstruct the input pixel value of each pixel in the encoded image block, based on the set of signaling bits.
    Type: Grant
    Filed: January 3, 2019
    Date of Patent: May 5, 2020
    Assignee: SONY CORPORATION
    Inventors: Toshinori Ihara, Mohammed Golam Sarwer, Ali Tabatabai
  • Patent number: 10606815
    Abstract: The present invention may be a method, system, and a computer program product. An embodiment of the present invention provides a method for creating one or more indexes for information retrieval. The method comprises reading a document having hinting information into memory; the hinting information being associated with each unique expression in an original document; and creating the indexes from the document which was read into memory. A first analysis method for generating a contiguous sequence of items from text in the document is used for creating the indexes for each sequence in the unique expression with which the hinting information is associated. A second analysis method for dividing the text into meaningful units is used for creating the indexes for each word in the text other than the unique expression. The first analysis method may be an n-gram analysis and the second analysis method may be a morphological analysis.
    Type: Grant
    Filed: March 29, 2016
    Date of Patent: March 31, 2020
    Assignee: International Business Machines Corporation
    Inventors: Hidekazu Fujiwara, Yoko Nameki, Soh Ohta
  • Patent number: 10567800
    Abstract: Techniques are described for performing transformation on video data. A transform circuit may receive M sample values of the video data from a pre-transform buffer, and process the M sample values with N computation units of the transform circuit to generate intermediate values. Processing the M sample values to generate the intermediate values includes feeding back temporary values from output of one or more of the N computation units to input of one or more of the N computation units. The transform circuit may store a first set of the intermediate values in a transpose buffer, and store a second set of the intermediate values in the pre-transform buffer that are to be later retrieved for storage in the transpose buffer.
    Type: Grant
    Filed: March 7, 2017
    Date of Patent: February 18, 2020
    Assignee: Qualcomm Incorporated
    Inventors: Yunqing Chen, Srikanth Alaparthi, Tushar Singhal, Harikrishna Reddy, Ashish Mishra
  • Patent number: 10542256
    Abstract: Embodiments of the present invention provide a method and a device for determining a transform block size, and can improve image compression efficiency. The method includes: obtaining, according to image block information and a splitting manner of an image block, a parameter of a transform block corresponding to the image block; and obtaining a transform block size according to the parameter of the transform block and the splitting manner of the image block. In the embodiments of the present invention, during determination of a transform block size, the transform block size is obtained according to a splitting manner of an image block and a parameter of a transform block corresponding to the image block. Therefore, a transform block matching the splitting manner of the image block can be used, improving image compression efficiency.
    Type: Grant
    Filed: July 21, 2017
    Date of Patent: January 21, 2020
    Assignees: Huawei Technologies Co., Ltd., Tsinghua University
    Inventors: Yuan Yuan, Xiaozhen Zheng, Yun He
  • Patent number: 10515319
    Abstract: A method for computing a probability that an object comprises a target includes: performing a scan of an area comprising the object, generating points; creating a segment corresponding to the object using the points as segment points, the segment extending from a first segment point to a last segment point, the segment comprising a plurality of the segment points; and applying a metric, computing the probability that the object comprises the target.
    Type: Grant
    Filed: December 16, 2016
    Date of Patent: December 24, 2019
    Assignee: Fetch Robotics, Inc.
    Inventors: Alex Henning, Michael Ferguson, Melonee Wise
  • Patent number: 10452651
    Abstract: Approaches for indexing and comparing charts are described. A system can receive one or more charts, which may include depictions of signals, and index portions of a chart using a sliding window algorithm. Subsequently, a system can receive a query that can be compared to the indexed portions of one or more charts. After a comparison, the most similar portions of the compared charts are provided based on a nearest neighbor search.
    Type: Grant
    Filed: December 23, 2014
    Date of Patent: October 22, 2019
    Assignee: Palantir Technologies Inc.
    Inventor: Guillem Palou
  • Patent number: 10410079
    Abstract: Wavelet transformation is performed on first image data and second image data until a decomposition level becomes a decomposition level based on synthesis control data or the like, and first wavelet coefficient data and second wavelet coefficient data are thereby generated. An ROI coefficient related to an ROI and a non-ROI coefficient in the first wavelet coefficient data are determined on the basis of mask data and the ROI coefficient in the first wavelet coefficient data and a wavelet coefficient in the second wavelet coefficient data are synthesized with each other, and synthesized coefficient data are thereby generated. Inverse wavelet transformation is performed on the synthesized coefficient data until a decomposition level becomes a predetermined end level, and synthetic image data are thereby generated.
    Type: Grant
    Filed: September 22, 2017
    Date of Patent: September 10, 2019
    Assignee: MegaChips Corporation
    Inventor: Yusuke Mizuno
  • Patent number: 10366472
    Abstract: Systems and methods in accordance with embodiments of the invention are disclosed that use super-resolution (SR) processes to use information from a plurality of low resolution (LR) images captured by an array camera to produce a synthesized higher resolution image. One embodiment includes obtaining input images using the plurality of imagers, using a microprocessor to determine an initial estimate of at least a portion of a high resolution image using a plurality of pixels from the input images, and using a microprocessor to determine a high resolution image that when mapped through the forward imaging transformation matches the input images to within at least one predetermined criterion using the initial estimate of at least a portion of the high resolution image.
    Type: Grant
    Filed: June 1, 2016
    Date of Patent: July 30, 2019
    Assignee: FotoNation Limited
    Inventors: Dan Lelescu, Gabriel Molina, Kartik Venkataraman
  • Patent number: 10237555
    Abstract: Methods, apparatus and systems for processing video blocks. The method including transforming, via a transform unit, the video residual data using a single one-dimensional transform to generate a first set of coefficients associated with the video block; quantizing, via a quantization unit, the first set of coefficients to generate a second set of quantized coefficients, wherein the method for processing the video residual data includes bit-shifting any of the first and second sets of coefficients; entropy coding the second set of quantized coefficients after the bit-shifting; and transmitting the entropy coded second set of quantized coefficients. In certain embodiments, the bit shifting may include integer and/or fractional bit shifting.
    Type: Grant
    Filed: July 12, 2017
    Date of Patent: March 19, 2019
    Assignee: VID Scale, Inc.
    Inventors: Jie Dong, Yan Ye
  • Patent number: 10205922
    Abstract: A display control apparatus which projects and displays at least one image on a screen, converts a frame frequency for a display of the at least one image to N times (N>1), generates an added image by adding a predetermined pattern to the converted frame and a subtracted image by subtracting the predetermined pattern from the converted frame as a first frame and a second frame following the first frame, projects the generated first frame and the second frame for an image on the screen at the N-times frequency, acquires the image projected on the screen, extracts the pattern from the acquired image, and corrects a shape of the image based on the extracted pattern.
    Type: Grant
    Filed: November 3, 2017
    Date of Patent: February 12, 2019
    Assignee: Canon Kabushiki Kaisha
    Inventor: Akihiro Ouchi
  • Patent number: 10110832
    Abstract: A system according to an exemplary embodiment includes: a first scaler configured to receive a plurality of touch events and down-scale the touch events to generate down-scaled touch events; a touch path logic unit configured to generate an output based on the down-scaled touch events; and a rendering logic unit configured to receive a video image, receive an output of the touch path logic unit, generate a display image by combining the video image with an overlay data according to the output of the touch path logic unit, and output the display image.
    Type: Grant
    Filed: May 12, 2015
    Date of Patent: October 23, 2018
    Assignee: Samsung Display Co., Ltd.
    Inventors: Bo Young An, Ho Suk Maeng, Weon Jun Choe
  • Patent number: 10110769
    Abstract: A computer implemented system and method for managing a stack containing a plurality of documents. The system scans and manages documents provided by the users in form of batches. Multiple users can provide the documents to be managed in form of a stack that contains the documents separated by separating pages and submission forms. The submission forms are then identified by the system to identify the batches and allot track numbers to the identified batches for future reference. Documents within the batches are identified by the separating pages and are allotted barcodes for identification. These documents are scanned and processed to obtain quality checked images of the documents which are then stored in a central repository. The system allows the users to change/set prioritization of a request or document type and also allows automatic indexing, routing of the transactions, processing, quality checking, and modification in the scanned images.
    Type: Grant
    Filed: December 23, 2014
    Date of Patent: October 23, 2018
    Assignee: TATA CONSULTANCY SERVICES LTD.
    Inventors: Vithal Satya Naga Prabhakara Medicherla, Suresh Babu Ponnan, Ohm Kumar Balasubramanian
  • Patent number: 9872026
    Abstract: Techniques related to video coding with sample adaptive offset coding are discussed. Such techniques may include setting a sample adaptive offset coding flag for a picture of a group of pictures based at least in part on a comparison of an available coding bit limit of the picture to a first threshold and a quantization parameter of the picture to a second threshold. In some examples, such techniques may also include setting the sample adaptive offset coding flag based on a coding structure associated with coding the group of pictures.
    Type: Grant
    Filed: June 12, 2015
    Date of Patent: January 16, 2018
    Assignee: Intel Corporation
    Inventors: Ximin Zhang, Sang-Hee Lee, Dmitry E. Ryzhov
  • Patent number: 9747001
    Abstract: An apparatus, that can be in the form of a wearable device, comprising a display, a memory, and a processor coupled to the display and memory. The processor retrieves a first set of points describing a first vector for a first display element; retrieves a second set of points describing a second vector for a second display element; modifies the first set of points by modifying one or more points in the first set of points to form a modification of the first vector; modifies the second set of points by modifying one or more points in the second set of points to form a modification of the second vector; and sequentially renders in the display the first vector, the modification to the first vector, the second vector with the modification to the second vector, and the second vector without the modification to the second vector.
    Type: Grant
    Filed: October 20, 2015
    Date of Patent: August 29, 2017
    Assignee: Pebble Technology Corp.
    Inventors: Heiko Behrens, Liron Mordechi Damir
  • Patent number: 9734588
    Abstract: A method and system is provided for the compression of a video captured from a static camera. A background image and an original image is captured from the static camera. A foreground object is extracted by taking the difference between the transformed original image and the transformed background image. The foreground object is represented as a sparse vector using wavelets. The method revolves around the compressive sensing framework by ingeniously using the complex field BCH codes and the syndrome as measurements to achieve robust background subtraction using reduced number of measurements. The reconstruction is carried out by a Complex-field BCH decoder coupled with block based implementation. According to another embodiment, an adaptive techniques has also been proposed for acquiring the measurements required for recovering the images.
    Type: Grant
    Filed: March 14, 2016
    Date of Patent: August 15, 2017
    Assignee: Tata Consultancy Services Limited
    Inventors: Narendra Narsimharao, Mariswamy Girish Chandra
  • Patent number: 9685056
    Abstract: A system and method for detecting change in a video stream may include receiving, by a processor, a video image from a video stream and a reference image. The processor may divide the video image and the reference image into non-overlapping blocks, and each block in the video image may correspond to a block in the reference image. The processor may perform an orthogonal transformation on a corresponding pair of blocks in the video image and the reference image. DC (direct current, or, average value) components in the orthogonally transformed pair of blocks in the video image and the reference image may be zeroed. An infinity-norm difference between coefficient values of the corresponding pair of transformed and modified blocks in the video image and the reference image may be determined and compared to a threshold.
    Type: Grant
    Filed: October 2, 2014
    Date of Patent: June 20, 2017
    Assignee: Qognify Ltd.
    Inventor: Gideon J. Klimer
  • Patent number: 9509997
    Abstract: A multi-viewpoint image which has been captured with respect to an object from a plurality of viewpoints is coded by setting a base viewpoint among the plurality of viewpoints. An image captured from a viewpoint that is not the base viewpoint is then coded using a reference image selected based on imaging parameters and a parallax.
    Type: Grant
    Filed: June 11, 2013
    Date of Patent: November 29, 2016
    Assignee: Canon Kabushiki Kaisha
    Inventors: Reiko Fujino, Satoshi Naito
  • Patent number: 9426467
    Abstract: An image decoding method which includes decoding transform coefficients for a target image. A combination of a vertical inverse transform matrix and a horizontal inverse transform matrix corresponding to a target image is set. The transform coefficients are inversely transformed along a vertical direction and a horizontal direction using the combination which has been set to obtain a prediction error. The decoded image is generated using the prediction error.
    Type: Grant
    Filed: February 21, 2014
    Date of Patent: August 23, 2016
    Assignee: KABUSHIKI KAISHA TOSHIBA
    Inventors: Jun Yamaguchi, Akiyuki Tanizawa
  • Patent number: 9424339
    Abstract: Embodiments of the invention form an information set from the current set of index information available by the operations of the Pre-Search and RunTime Search components of the search engine. A search request that contains search terms and/or other search criteria (e.g. date or file type) is entered by a user through an input interface. The search terms and the information set are worked through the search engine modules to provide the actual results sought by the user. These results are provided to the user via a Return Interface. Embodiments involve a search engine that forms a numeric code index that includes grammar embedded rules from a plurality of documents, converts a search request into the same format as the index, and then compares the index and the search request to form the search results.
    Type: Grant
    Filed: August 15, 2008
    Date of Patent: August 23, 2016
    Inventors: Athena A. Smyros, Constantine Smyros
  • Patent number: 9391635
    Abstract: A block encode circuit (800) including a scanner (820) operable to scan a block having data values spaced apart in the block by run-lengths to produce a succession of pairs of values of Level and Run representing each data value and run-length, and wherein the Level values include one or more AC values succeeded by a DC value in the succession, and a Run-Level encoder (830) responsive to said scanner (820) to encode the values of Level and Run in a same AC to DC order as in the succession of pairs of values from said scanner (820) to deliver an encoded output. Other encoders, decoders, codecs and systems and processes for their operation and manufacture are disclosed.
    Type: Grant
    Filed: May 10, 2010
    Date of Patent: July 12, 2016
    Assignee: Texas Instruments Incorporated
    Inventor: Yusuke Minagawa
  • Patent number: 9342901
    Abstract: A method for generating a shader that is used by a rendering engine to render a visual representation of a computer model. A shader generation engine receives a set of surface data that describes a surface in view of various lighting conditions. The shader generation engine compresses the set of surface data to generate a compressed representation of the set of surface data based on a selected compression algorithm. The shader generation engine generates a shader based on the compressed representation that is configured to be implemented with a rendering engine, and generates a set of shader data based on the compressed representation that includes a set of material characteristics for coloring pixels of the visual representation. Advantageously, the shader generation process is simplified because different compression algorithm-rendering engine shader combinations can be generated without manually programming the shaders.
    Type: Grant
    Filed: October 27, 2008
    Date of Patent: May 17, 2016
    Assignee: AUTODESK, Inc.
    Inventors: Jérôme Maillot, Eric Bourque
  • Patent number: 9342749
    Abstract: Systems, apparatus, articles, and methods are described related to a hardware-based convolution pre-filter to accelerate object detection.
    Type: Grant
    Filed: December 18, 2012
    Date of Patent: May 17, 2016
    Assignee: Intel Corporation
    Inventors: Jianguo Li, Ya-Ti Peng, Yi-Jen Chiu
  • Patent number: 9330372
    Abstract: Disclosed herein are system, method, and computer program product embodiments for generating an optimal development infrastructure. An embodiment operates by decomposing a development process into one or more functional requirements, mapping each functional requirement of the decomposed development process to the one or more corresponding functional capabilities, and selecting one or more services operable to implement each mapped functional capability, wherein each service comprises one or more functional capabilities.
    Type: Grant
    Filed: June 28, 2013
    Date of Patent: May 3, 2016
    Assignee: SAP AG
    Inventors: Oleg Koutyrine, Klaus Kopecz
  • Patent number: 9332277
    Abstract: An invention is disclosed for performing differencing of graphical data in post-transform space for a remote presentation session. Graphical data is transformed from a first representation to a second representation (e.g. with a DWT), and then a difference is taken of the post-transform data and the post-transform data of the frame that preceded the current frame. This difference is then encoded and transmitted to a client, which decodes it, and creates a representation of the graphical data using the delta, and a previously determined representation of the previous frame. By performing differencing in post-transform space, fidelity of the remote presentation session is retained while it may decrease bandwidth. This may occur because the entropy of the delta representation is usually lower than a non-delta representation while the scheme retains the identical data of the final decoded image of the non-delta version of the same compression scheme.
    Type: Grant
    Filed: March 9, 2015
    Date of Patent: May 3, 2016
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Paul Johan Hsieh, Denis Sergeyevich Shchepetov, Sridhar Sankuratri, Damien Gerard Saint Macary
  • Patent number: 9253495
    Abstract: Methods and apparatus are provided for improved chroma transforms for inter frames in video encoding and decoding. The apparatus and method encode at least a portion of a picture using inter-coding, wherein a transform is adaptively selected, from among a plurality of transforms, to apply to chroma components of the portion responsive to at least one selection criterion.
    Type: Grant
    Filed: November 24, 2010
    Date of Patent: February 2, 2016
    Assignee: THOMSON LICENSING
    Inventors: Xiaoan Lu, Peng Yin, Qian Xu, Joel Sole, Yunfei Zheng
  • Patent number: 9232236
    Abstract: A video coding method enables reduction in the decrease in the efficiency of coding a video and enhancement of the image quality of the resulting decoded video. The video coding method includes predicting a video signal to generate a prediction signal; calculating, as a prediction error signal, the difference between the video signal and a prediction signal; and generating a reconstructed video signal by reconstructing the video signal based on the prediction signal and the prediction error signal. The video coding method also includes determining filter data items that are used to filter each of at least two of the prediction signal, the prediction error signal, and the reconstructed video signal; and coding the determined filter data items based on the cross correlations between the determined filter data items.
    Type: Grant
    Filed: April 2, 2010
    Date of Patent: January 5, 2016
    Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICA
    Inventors: Matthias Narroschke, Thomas Wedi, Steffen Wittmann
  • Patent number: 9191668
    Abstract: There is described an image compressing/decompressing method and device that provides the lossless data compression/decompression scheme in two separate modules. Partially entropy encoded/decoded data is written to memory by a first module after a first part of the process and retrieved by a second module to perform the second part of the lossless compression/decompression scheme.
    Type: Grant
    Filed: October 7, 2014
    Date of Patent: November 17, 2015
    Assignee: MATROX GRAPHICS INC.
    Inventors: Jean-Jacques Ostiguy, Simon Garneau, Oui Goi
  • Patent number: 9129405
    Abstract: An image processing apparatus generates an image appended with additional effective pixels by executing appending processing for appending additional effective pixels by extending input image data in a scanning direction using pixels at terminal ends of the input image data in the scanning direction. The image processing apparatus generates image data of a second hierarchy obtained by reducing the input image data by applying filtering to the generated image appended with the additional effective pixels, and generates image data of a third hierarchy by applying filtering to the image data of the second hierarchy.
    Type: Grant
    Filed: January 28, 2013
    Date of Patent: September 8, 2015
    Assignee: CANON KABUSHIKI KAISHA
    Inventor: Shutaro Kunieda
  • Patent number: 9123089
    Abstract: Techniques and tools for signaling and using image tiling information (such as syntax elements relating index tables and header size), signaling and using windowing information (such as techniques for using windowing parameters when rotating, cropping or flipping images), and signaling and using alpha channel information are described.
    Type: Grant
    Filed: August 19, 2013
    Date of Patent: September 1, 2015
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Sridhar Srinivasan, Chengjie Tu, Shankar Regunathan
  • Patent number: 9118923
    Abstract: Presented herein are reconstruction and overlap transform system(s) and method(s) supporting VC-1 decoding. In one embodiment, there is presented a system for reconstructing pixels. The system comprises a first circuit and a second circuit. The first circuit reconstructs pixels from residuals, where the residuals encode video data in accordance with a first encoding standard. The second circuit reconstructs pixels from residuals, where the residuals encode video data in accordance with a second encoding standard.
    Type: Grant
    Filed: April 27, 2006
    Date of Patent: August 25, 2015
    Assignee: Broadcom Corporation
    Inventor: Stephen Gordon
  • Patent number: 9113143
    Abstract: Techniques are disclosed for detecting an out-of-focus camera in a video analytics system. In one embodiment, a preprocessor component performs a pyramid image decomposition on a video frame captured by a camera. The preprocessor further determines sharp edge areas, candidate blurry edge areas, and actual blurry edge areas, in each level of the pyramid image decomposition. Based on the sharp edge areas, the candidate blurry edge areas, and actual blurry edge areas, the preprocessor determines a sharpness value and a blurriness value which indicate the overall sharpness and blurriness of the video frame, respectively. Based on the sharpness value and the blurriness value, the preprocessor further determines whether the video frame is out-of-focus and whether to send the video frame to components of a computer vision engine and/or a machine learning engine.
    Type: Grant
    Filed: June 28, 2013
    Date of Patent: August 18, 2015
    Assignee: BEHAVIORAL RECOGNITION SYSTEMS, INC.
    Inventors: Ming-Jung Seow, Dennis G. Urech
  • Patent number: 9105074
    Abstract: A global bit stream comprising at least a first and a second bit stream is generated using an original visual bit stream. The first bit stream is generated by compressing the original visual bit stream using a JPEG-LS compression technique with a lossy factor ‘n’. The second bit stream is generated by compressing a difference bit stream using a JPEG-LS lossless compression technique, wherein the difference bit stream comprises a difference between the original visual bit stream and the first bit stream. The visual display is recreated using a reconstructed bit stream after receiving the global bit stream over a channel, wherein the reconstructed bit stream is generated by decompressing at least the first bit stream using a JPEG-LS lossless compression technique. The second bit stream is decompressed using JPEG-LS lossless compression technique and the quality of the display may be enhanced by adding the decompressed second bit stream.
    Type: Grant
    Filed: October 27, 2009
    Date of Patent: August 11, 2015
    Assignee: Intel Corporation
    Inventors: Andrey Mikhailovich Turlikov, Anton Valerievich Sergeev, Anna Sergeyevna Ukhanova
  • Patent number: 9055251
    Abstract: Methods and systems divide an image into image blocks, determine a number of high-variation blocks within valid image blocks, determine a page variance measure by dividing the number of high-variation blocks by the total number of valid image blocks, and classify the image as comprising a halftone and/or text image if the page variance measure exceeds a halftone threshold. Such methods and systems classify the image as being a continuous tone image if the page variance measure does not exceed the halftone threshold and a text index is below a text index threshold, and classify the image as being a continuous tone and text image if the page variance measure does not exceed the halftone threshold and the text index is not below the text index threshold. Further, such methods and systems select an image processing mode based on such image classifications.
    Type: Grant
    Filed: September 24, 2013
    Date of Patent: June 9, 2015
    Assignee: Xerox Corporation
    Inventors: Dhevendra Alagan Palanivel, Sainarayanan Gopalakrishnan, Xing Li, Clara Cuciurean-Zapan
  • Patent number: 9055338
    Abstract: A video coding method comprises dividing a coefficient vector into two or more sub-vectors, each sub-vector having a length less than or equal to a length of the coefficient vector, indicating a size of each sub-vector and a position of each sub-vector within the coefficient vector, and forming a block corresponding to each sub-vector.
    Type: Grant
    Filed: March 12, 2008
    Date of Patent: June 9, 2015
    Assignee: Nokia Corporation
    Inventors: Justin Ridge, Xianglin Wang
  • Patent number: 9031341
    Abstract: In an image coding apparatus (1), a Hadamard transform unit (11) performs horizontal Hadamard transform on a picture of uncompressed image data (21). The sum total of absolute values of AC component values obtained by the Hadamard transform is calculated as a Hadamard value (23) of the picture. A scene change determination unit (12) determines whether a scene change occurs or not in the picture on the basis of the Hadamard value (23). In a case where a scene change occurs in the picture or where a differential absolute value between the amount of generated codes in a coded GOP and the ideal amount of codes in a GOP is larger than a predetermined reference value, a quantization parameter determination unit (13) determines a quantization parameter (24) of the picture on the basis of the Hadamard value (23) and the target amount of codes of the picture. A coding unit (14) codes the picture by using the determined quantization parameter (24).
    Type: Grant
    Filed: February 24, 2012
    Date of Patent: May 12, 2015
    Assignee: MegaChips Corporation
    Inventors: Naotsugu Yamamura, Akira Okamoto, Makoto Saito, Takuya Hiraoka
  • Publication number: 20150125087
    Abstract: A method for compressing an image including one or more blocks, each of the one or more blocks including an x channel, a y channel, and a z channel, the method including, for each of the one or more blocks: performing a frequency domain transform, by a processor, on each of the x, y, and z channels; ordering, by the processor, the transformed x, y, and z channels into first, second, and third transformed channels; performing an inter-channel transform in accordance with the order of the first transformed channel, the second transformed channel, and the third transformed channel; and encoding the inter-channel transformed channels.
    Type: Application
    Filed: October 27, 2014
    Publication date: May 7, 2015
    Inventor: Dihong Tian
  • Patent number: 9025900
    Abstract: A signal processing transformation (wavelet, Fourier, discrete cosine) is applied to a digital image on a mobile device in order to produce a low-level information image and at least one high-level information image. The low-level image is recognizable as the digital image and is kept on the device; all other related images are deleted. The high-level information images are uploaded. The transformation is applied recursively and is dictated by a default setting, calculated from data of the mobile device, or input. To regenerate the original image the device connects to the server and downloads a set of high-level information images or all sets. The low resolution image is combined with the high-level information images using the reverse of the transformation originally applied to produce a higher resolution version of the low resolution image. Successive sets of high-level information images may be recursively applied to generate successively higher resolution images.
    Type: Grant
    Filed: September 7, 2012
    Date of Patent: May 5, 2015
    Assignee: Trend Micro Inc.
    Inventors: Wen-Kwang Tsao, Kuan-Ru Fu, Yu-Kuang Lin
  • Patent number: 9020290
    Abstract: A method for removing boundary distortion includes receiving a one-dimensional input signal and determining whether the one-dimensional input signal includes an even number of data elements. If the one-dimensional input signal includes an even number of data elements, an extrapolation operation is performed on one of a first or second boundary of the one-dimensional input signal. The extrapolation operation produces one additional data element. The method may further include performing a mirroring operation on the signal data resulting from the extrapolation operation to produce a mirrored signal, and performing a transform operation the mirrored signal.
    Type: Grant
    Filed: January 30, 2013
    Date of Patent: April 28, 2015
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Damien Saint Macary, Sridhar Sankuratri
  • Patent number: 9008451
    Abstract: An image encoding and/or decoding apparatus and method are provided. The image encoding apparatus includes: a transform unit removing spatial redundancy by transforming an original image being input; an allowable noise obtaining unit obtaining an allowable noise from the original image; a quantization parameter determination unit determining a quantization parameter by using the allowable noise; a quantization unit generating a quantized coefficient, by quantizing a transform coefficient provided from the transform unit by using the quantization parameter; and an entropy encoding unit entropy encoding the quantized coefficient to remove statistical redundancy.
    Type: Grant
    Filed: December 14, 2005
    Date of Patent: April 14, 2015
    Assignee: Samsung Electronics Co., Ltd.
    Inventors: Si-Hwa Lee, Doo-Hyun Kim, Hye-Yun Kim, In-So Kweon